Field of the Invention
The present invention relates to an image capturing apparatus and a control method of an image sensor capable of reading out from each pixel a pair of signals corresponding to light beams that have passed through different pupil regions of an optical system.
Description of the Related Art
Conventionally, a technique for realizing focus detection function and an image signal acquisition using one image sensor has been known, and as an example, an image sensor capable of using pixels for acquiring information for focus detection as pixels for acquiring an image for display/record has been proposed (see Japanese Patent Laid-Open No. 2007-325139). According to this technique, among pixels that form an image sensor, each of at least part of the pixels is divided into a plurality of regions in the horizontal and/or vertical direction, and upon acquiring an image, an image signal is obtained by adding signals obtained from the divided regions. Further, upon performing focus control, the divided regions are read out so that signals of two regions of each pixel that is pupil-divided in the horizontal or vertical direction are obtained, and the obtained signals can be used for focus detection of phase difference method.
Further, Japanese Patent Laid-Open No. 2012-155095 proposes an image capturing apparatus capable of changing between a control of outputting signals of divided areas as an image signal by adding the signals by pixel, and a control of outputting the signals of the divided areas independently so as to be used as focus control signals for phase difference method.
On the other hand, as the number of pixels in an image sensor increases, it is necessary to readout larger number of pixels than ever within a predetermined period of time, which causes an increase in system load due to an increase in output rate of the image sensor or an increase in the number of output channels. For example, there is a problem in which, if the number of output channels increases, or if the output rate increases with the number of channels being unchanged, adjustment of signal delay amount between the channels becomes more complicated. Especially, in a case of image sensor having pupil division readout function, in which each pixel is configured with a plurality of photoelectric conversion portions, the number of signals to be read out increases, and the load of system further increases. Further, a problem arises such that energy consumption increases by reading a larger number of signals.
According to Japanese Patent Laid-Open No. 2012-155095, in a focus control area, a signal read out from each of the photoelectric conversion portions is output as-is from the image sensor, and in other areas, signals read out from the photoelectric conversion portions are added by pixel and output from the image sensor. By using different output methods in the focus control area and in the other areas, the number of signals to be output is reduced. However, in Japanese Patent Laid-Open No. 2012-155095, since a signal is independently read out from each of the photoelectric conversion portions, the period of time required for reading out signals cannot be sufficiently shortened.
Further, according to Japanese Patent Laid-Open No. 2012-155095, the focus control region is set to an object region, and a way of outputting the signal for focus control and a way of outputting a signal for image are changed. In this case, there is a problem in which, when a subject is moving, it is difficult to perform focus control while sufficiently tracking the moving subject. The accuracy of performing focus detection while tracing a moving subject improves as the number of regions from which signals are read out without being added increases, but the problems regarding the increase in system load and the increase in power consumption cannot be solved. Therefore, a problem that from which area in a frame a focus control signal of phase difference method is to be read out is posed.